1. Field of the Invention
The present invention relates to a moving-image reproducing apparatus, a moving-image reproducing method, a program, and a recording medium for decompressing a digital compressed moving image to reproduce the decompressed moving image.
2. Description of the Related Art
Along with development of technologies for efficiently encoding and compressing digital videos in recent years, not only personal computers (PCs) but also digital cameras and mobile phones are capable of decompressing and decoding digital videos that are encoded and compressed to reproduce the digital videos.
In typical encoding and compression methods for digital videos, for example, in H.264 and Moving Picture Experts Group (MPEG), each image (hereinafter referred to as a frame) composing a video is divided into small rectangular blocks (hereinafter referred to as macroblocks (MBs)). Space correlation or time correlation is used for each MB to generate predicted data. Then, after a transformation coefficient (difference coefficient) resulting from orthogonal transformation of the difference between the image data about the MB and the predicted data is subjected to quantization and scan conversion, variable length coding is performed and space redundancy is reduced to achieve a higher coding efficiency.
However, it is necessary to make a large amount of calculation and perform memory transfer in order to reproduce such a digital video that is subjected to encoding and compression. Accordingly, restrictions on resolution, frame rate, bit rate, etc. often apply to reproducible video data from the viewpoint of the processing capacity and the power consumption.
In such a situation, moving-image reproducing apparatuses capable of decreasing the amount of memory transfer and reducing the power consumption are proposed. For example, Japanese Patent Laid-Open No. 2000-115777 and Japanese Patent Laid-Open No. 2007-67526 disclose technologies in which a prediction frame memory and/or a display frame memory is not updated if an input bit stream is decoded and an MB to be decompressed satisfies a certain condition on the basis of encoding information that is output in order to reduce the amount of memory transfer.
In addition, it is possible to capture still images during recording of moving images in recent years. In recording of still images during capturing of moving images, a moving image file can be created into which blackout images and/or alternate frames for the still images are inserted so that the moving image file is not divided into two segments with the capturing of the still image sandwiched therebetween while the recording of the moving image is interrupted by the capturing of the still image. An image pickup apparatus capable of decreasing the data size of moving image data to reduce the load of writing on a recording medium is disclosed (Japanese Patent Application No. 2008-053415 (Japanese Patent Laid-Open No. 2008-301475)). The image pickup apparatus disclosed in Japanese Patent Application No. 2008-053415 generates moving image data shown in FIG. 6. Referring to FIG. 6, the blackout images for multiple frames and the still-image alternate images (stop motion images) for multiple frames, which are inserted while recording of moving images is interrupted by capturing of still images, are represented as skipped macroblock images (hereinafter referred to as SMB images). In each SMB image, all the MBs are SMBs. The SMB image is one kind of an interframe compressed image and indicates information about the difference between the frames. For example, the SMB image has no motion vector. The SMB image is image data having a very small data size on the order of several tens of bytes even in Video Graphics Array (VGA). The SMB image represents that completely the same image as that displayed in the previous frame is displayed.
However, it is necessary to refer to the previous frame for every MB in order to reproduce the SMB image in reproduction of the moving image data shown in FIG. 6. Accordingly, the amount of memory transfer is greatly increased to cause an increase in processing load and power consumption. Consequently, it is desirable to reduce the amount of memory transfer and the power consumption.